CN102759573A - Frequency change-based structure damage positioning and damage degree evaluating method - Google Patents

Abstract

A frequency change-based structure damage positioning and damage degree evaluating method is characterized by comprising the following steps: establishing a finite element model of a non-damage structure as a reference finite element model; carrying out vibration test on a damage structure, and recognizing m-order modal frequency of the damage structure; carrying out damage degree evaluation on the damage structure; and confirming the damaged position of the damage structure. The method can accurately position the damage and evaluate the damage degree only by utilizing finite low-order frequency information before and after the damage of the structure without the need of structure vibration mode information, and is applicable to single damage and multiple damage working conditions, thus having certain practical application values.

Description

Based on the structural damage location of change of frequency and the appraisal procedure of degree of injury

Technical field

The present invention relates to the method for structural damage location and degree of injury assessment; Relate in particular to a kind of low order frequency that can utilize damaged structure to survey and confirm the damage position of structure, and the structural damage based on change of frequency of the order of severity of assessment damage is located and the appraisal procedure of degree of injury.Belong to marine oil engineering field.

Background technology

It is inevitably that heavy construction structures such as ocean platform are damaged at it during one's term of military service, and guarantees that the unique method of structural safety is a damage position of determining structure as early as possible, and the degree of damage is assessed, so that carry out maintenance and reinforcement timely.

At present, the detection method for structural damage mainly comprises: local lossless detection methods such as range estimation and ultrasonic, magnetic, acoustic emission etc.Yet, owing to more weak visual observation condition and damage location might have been hindered the effect of range estimation by reasons such as marine growth coverings.In addition, in the said method damage field of technical requirement structure be with known as condition precedent, require to be equipped with special extra testing apparatus and professional.Therefore, said method is not too convenient to the detection of ocean platform, and testing cost is expensive.

Comparing with said method, is simple relatively, lower-cost based on the diagnosing structural damage technology of vibration-testing, is acknowledged as more promising damage detecting method of overall importance.The ultimate principle of this method is: damage will cause the system stiffness of structure to change; Thereby cause the variation of dynamic Characteristics of Structure parameter (like the frequency of structure, vibration shape etc.); So; The dynamic Characteristics of Structure parameter can be as the index of diagnosing structural damage, and whether be used for decision structure has damage to take place, and and then the order of severity of assessment damage.The advantage that these class methods are the most outstanding is to utilize the dynamic response test under the environmental excitation to carry out damage, and whole damage operating process can not influence the operate as normal of structure, can accomplish the online detection and the assessment of structural damage easily.

The damage of structure comprises four levels:

(1) judges whether structure is damaged (damage identification);

(2) confirm the damage position (damage location) of structure;

(3) degree of injury of evaluation structure (degree of injury assessment);

(4) prediction of structure residual life.

Based on the damage diagnosis method process years development that structural dynamic characteristic changes, the researchist proposes also to have developed many methods, but mostly concentrates on one, two levels.The achievement in research that can damage the assessment of location and degree of injury simultaneously is less, and these methods especially are applied to large scale structure and exist certain limitation following in practical application:

(1) needs the modal information of high-order: because aspects such as the frequency range of exciting load and recognition technology generally are difficult to obtain high order mode information;

(2) need complete modal information: because the restriction of number of sensors is difficult to measure with some degree of freedom information (like rotational freedom), the modal parameters that actual test obtains is incomplete;

(3) need the normalized formation information of quality: only utilize the output response to carry out Modal Parameter Identification down to the environmental load excitation, the vibration shape that its identification obtains can't quality normalization.

How to overcome these shortcomings, the limited lower mode information of only utilizing identification to obtain is carried out damage check, is the emphasis and the difficult point of research at present.

Summary of the invention

Fundamental purpose of the present invention is to overcome the above-mentioned shortcoming that prior art exists; And provide a kind of based on the structural damage location of change of frequency and the appraisal procedure of degree of injury; It does not need the vibration shape information of structure, only utilizes the limited low order frequency information of structural damage front and back can damage location and degree of injury assessment exactly; Be applicable to single injury, multiple damage operating mode, have certain actual application value.

The objective of the invention is to realize by following technical scheme:

A kind of based on the structural damage location of change of frequency and the appraisal procedure of degree of injury, it is characterized in that: adopt following steps:

The first step: set up the finite element model of last damaged structure, as the benchmark finite element model;

Second step: damaged structure is carried out vibration-testing, the m rank model frequency of identification of damage structure ${\mathrm{\ω}}_i^{*}(i=1,...,m);$

The 3rd step: damaged structure is carried out the degree of injury assessment;

The 4th step: damaged structure is carried out damage position confirm.

In the said first step, the benchmark finite element model representes with K and M, that is: the not stiffness matrix of damaged structure and mass matrix, and its list of feature values is shown

$K{\mathrm{\Φ}}_i={\mathrm{\ω}}_i^{2}M{\mathrm{\Φ}}_i---\left(1\right)$

Wherein, ω _iAnd Φ _iBe respectively the i order frequency and the vibration shape of benchmark finite element model.

In said second step, the m rank model frequency of identification of damage structure

concrete steps are following:

(1), utilize the structural dynamic response data after the sensor measurement works damage, and data storage is gone in the storer;

(2), from storer, read stored parameters information, utilize its limited lower mode frequency of Modal Parameters Identification identification as damaged structure model frequency

Said damaged structure is with the not stiffness matrix and the mass matrix eigenwert formula of damaged structure:

Be benchmark, with K ^*And M ^*The integral rigidity matrix and the total quality matrix of expression damaged structure, structural damage generally only causes the variation of the rigidity of structure, and very little to the quality influence of structure, sets M ^*=M, then its eigenwert does

$K^{*}{\mathrm{\Φ}}_i^{*}={\mathrm{\ω}}_i^{*2}M{\mathrm{\Φ}}_i^{*}---\left(2\right)$

Wherein,

and

is respectively the i order frequency and the vibration shape of damaged structure; Utilize formula (1) and (2), obtain

${\left({\mathrm{\Φ}}_i\right)}^T{K}^{*}{\mathrm{\Φ}}_i^{*}=\frac{{\mathrm{\ω}}_i^{*2}}{{\mathrm{\ω}}_i^2}{\left({\mathrm{\Φ}}_i\right)}^{T}K{\mathrm{\Φ}}_i^{*}---\left(3\right)$

Wherein, subscript T represents transpose of a matrix, and promptly ranks exchange;

Setting structure has N _dDamage has taken place in individual unit, and damage position is known, and then the integral rigidity matrix representation of damaged structure does

$K^{*}=K+\underset{n=1}{\overset{N_d}{\mathrm{\Σ}}}{\mathrm{\α}}_n{K}_{l_n}---\left(4\right)$

Wherein, α _nAnd l _nBe respectively the degree of injury of n damage unit and the unit number of n damage unit; With formula (4) substitution formula (3), can get

$\underset{n=1}{\overset{N_d}{\mathrm{\Σ}}}{\mathrm{\α}}_n{\left({\mathrm{\Φ}}_i\right)}^T{K}_{l_n}{\mathrm{\Φ}}_i^{*}=\frac{{\mathrm{\ω}}_i^{*2}-{\mathrm{\ω}}_i^2}{{\mathrm{\ω}}_i^2}{\left({\mathrm{\Φ}}_i\right)}^{T}K{\mathrm{\Φ}}_i^{*}---\left(5\right)$

Definition $C_i={\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*},$ $C_{n,i}={\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}$ With $b_i=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="HDA00001929722200011.png,HDA00001929722200021.png"\; state="\{\{state\}\}">i</figure-callout>}}^2}{{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="HDA00001929722200011.png,HDA00001929722200021.png"\; state="\{\{state\}\}">i</figure-callout>}}^2},$ Then formula (5) can be written as

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{c}_{n,i}=b_i---\left(6\right)$

Write as matrix form, promptly

Cα＝b (7)

Set the preceding m order frequency information that has recorded damaged structure, then C is that dimension is m * N _dMatrix, α and b are respectively that dimension is N _d* 1 and the column vector of m * 1; Through least square method solution formula (7), obtain

α＝(C ^TC) ^-1C ^Tb (8)

α is the degree of injury of structural unit.

In said the 3rd step, the concrete steps of damaged structure being carried out the degree of injury assessment are following:

(1) the setting damage position is known; Utilize the model frequency

of benchmark finite element model and damaged structure to make up m equation, that is:

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2}{\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*}$

Wherein, K is the global stiffness matrix of benchmark finite element model;

And α _nBe respectively the element stiffness matrix and the degree of injury of n damage unit; N _dNumber for the damage unit; Φ _iWith

Be respectively the i first order mode of benchmark finite element model and damaged structure; ω _iWith Be respectively the i order frequency of benchmark finite element model and damaged structure; Subscript T represents transpose of a matrix;

Definition $C_i={\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*},$ $C_{n,i}={\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}$ With $b_i=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="HDA00001929722200011.png,HDA00001929722200021.png"\; state="\{\{state\}\}">i</figure-callout>}}^2}{{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="HDA00001929722200011.png,HDA00001929722200021.png"\; state="\{\{state\}\}">i</figure-callout>}}^2},$ Then following formula can be written as matrix form C α=b;

(2) iterative degree of injury;

The concrete calculation procedure of iterative degree of injury is:

1. iteration initial assignment, k=0, setting degree of injury is 0, i.e. α ⁽⁰⁾=0, subscript " 0 " is represented iterative initial value;

2. iteration begins, k=k+1, by $K^{*\left(k\right)}=K+\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n^{(k-1)}{K}_{l_n}$ And formula $K^{*\left(k\right)}{\mathrm{\&Phi;}}_i^{*\left(k\right)}={\mathrm{\&omega;}}_i^{*2}M{\mathrm{\&Phi;}}_i^{*\left(k\right)}$ Calculate

3. 2. utilization calculates Compute matrix C and vectorial b, last, by α=(C ^TC) ^-1C ^TThe b calculation of alpha ^(k)

4. set stopping criterion for iteration, if max{| is α ^(k)-α ^(k-1)|≤tol sets up, iteration stopping then, and wherein, tol is predefined allowable error, then α ^(k)Be the degree of injury estimated value; Continue otherwise return 2..

In said the 4th step, the concrete calculation procedure that damage position is confirmed is:

(1) sets total N _kIndividual damage operating mode is utilized N _mClass frequency changes carries out the degree of injury assessment to each damage operating mode, obtains N _mIndividual degree of injury estimated value

I=1 ..., N _m, subscript k representes k the damage operating mode of setting;

(2) utilize the degree of injury estimated value

Make up damage location factor D I _k, it is defined as

${\mathrm{DI}}_k=\frac{\min \left(e_k\right)}{e_k}$

In the following formula

Represent corresponding k N that sets the damage operating mode _mThe average of group estimated value, σ _kBe its standard deviation, min (e _k) represent e _kGet minimum value; Then to equal 1 unit be real damage position to the damage location factor.

The invention has the beneficial effects as follows:

(1) because it does not need the vibration shape information of structure, only utilize the variation of frequency to calculate, and frequency information obtains more easily than vibration shape information, therefore, avoided actual measurement vibration shape completeness, quality normalization and the lower shortcoming of accuracy of identification;

(2) only utilize the limited low order frequency information in damage front and back can carry out damage position and confirm and the degree of injury assessment that therefore, the accuracy of identification of frequency is higher than the accuracy of identification of the vibration shape, can reduce to the Modal Parameter Identification error minimum.

(3) be applicable to single injury, multiple damage operating mode, the damage position of structure can be confirmed exactly and assess its degree of injury to have certain actual application value.

Description of drawings:

Fig. 1 is a plane of the present invention steel frame construction finite element model synoptic diagram.

Fig. 2 is the damage operating mode synoptic diagram of plane of the present invention steel frame construction.

Fig. 3 is the damage locating effect figure (operating mode 1) of plane of the present invention steel frame construction.

Fig. 4 is the damage locating effect figure (operating mode 2) of plane of the present invention steel frame construction.

Fig. 5 is the damage locating effect figure (operating mode 3) of plane of the present invention steel frame construction.

Fig. 6 is the damage locating effect figure (operating mode 4) of plane of the present invention steel frame construction.

Fig. 7 is the damage locating effect figure (operating mode 5) of plane of the present invention steel frame construction.

Main label declaration among the figure:

Vertically bar unit, 2. the horizontal bars unit, 3. the brace unit, 4. the horizontal bars unit,

Vertically bar unit, 6. the horizontal bars unit, 7. the brace unit, 8. the horizontal bars unit,

Vertically bar element, 10. 11. brace unit, horizontal bars unit, 12. horizontal bars unit,

13. vertically bar unit, 14. horizontal bars unit, 15. brace unit, 16. horizontal bars unit,

17. vertically bar unit, 18. horizontal bars unit, 19. brace unit, 20. horizontal bars unit,

21. vertical bar unit.

Embodiment

Like Fig. 1, shown in Figure 2, the present invention adopts following steps:

The first step: set up the not finite element model of damaged structure, as the benchmark finite element model;

Second step: damaged structure is carried out vibration-testing, the m rank model frequency of identification of damage structure ${\mathrm{\&omega;}}_i^{*}(i=1,...,m);$

The 3rd step: damaged structure is carried out the degree of injury assessment;

The 4th step: damaged structure is carried out damage position confirm.

Above step specific algorithm is derived as follows:

One, represent the benchmark finite element model with K and M, that is: the not stiffness matrix of damaged structure and mass matrix, its list of feature values is shown

$K{\mathrm{\&Phi;}}_i={\mathrm{\&omega;}}_i^{2}M{\mathrm{\&Phi;}}_i---\left(1\right)$

Wherein, ω _iAnd Φ _iBe respectively the i order frequency and the vibration shape of benchmark finite element model;

Damaged structure is with the not stiffness matrix and the mass matrix eigenwert formula of damaged structure: Be benchmark, with K ^*And M ^*The integral rigidity matrix and the total quality matrix of expression damaged structure, the damage of structure generally only causes the variation of the rigidity of structure, and very little to the quality influence of structure, sets M ^*=M, then its eigenvalue problem does

$K^{*}{\mathrm{\&Phi;}}_i^{*}={\mathrm{\&omega;}}_i^{*2}M{\mathrm{\&Phi;}}_i^{*}---\left(2\right)$

Wherein,

and is respectively the i order frequency and the vibration shape of damaged structure.Utilize formula (1) and (2), can obtain

${\left({\mathrm{\&Phi;}}_i\right)}^T{K}^{*}{\mathrm{\&Phi;}}_i^{*}=\frac{{\mathrm{\&omega;}}_i^{*2}}{{\mathrm{\&omega;}}_i^2}{\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*}---\left(3\right)$

Wherein, subscript T represents transpose of a matrix, and promptly ranks exchange;

Setting structure has N _dDamage has taken place in individual unit, and damage position is known, and then the integral rigidity matrix of damaged structure can be expressed as

$K^{*}=K+\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{K}_{l_n}---\left(4\right)$

Wherein, α _nAnd l _nBe respectively the degree of injury of n damage unit and the unit number of n damage unit.With formula (4) substitution formula (3), can get

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2}{\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*}---\left(5\right)$

Definition $C_i={\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*},$ $C_{n,i}={\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}$ With $b_i=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2},$ Then formula (5) can be written as

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{c}_{n,i}=b_i---\left(6\right)$

Write as matrix form, promptly

Cα＝b (7)

Set the preceding m order frequency information that has recorded damaged structure, then C is that dimension is m * N _dMatrix, α and b are respectively that dimension is N _d* 1 and the column vector of m * 1.Through least square method solution formula (7), obtain

α＝(C ^TC) ^-1C ^Tb (8)

α is the degree of injury of structural unit.

Need to prove; Utilize formula (8) to find the solution damage of structure and need use the complete vibration shape information of damaged structure, this is impossible in practical application, in order to overcome this restrictive condition; Adopt iterative technique to come the degree of injury of computation structure, concrete steps are following:

The 1st step: iteration initial assignment; K=0, setting degree of injury is 0, i.e. α ⁽⁰⁾=0, subscript " 0 " is represented iterative initial value.

The 2nd step: iteration begins, k=k+1; By $K^{*\left(k\right)}=K+\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n^{(k-1)}{K}_{l_n}$ And formula $K^{*\left(k\right)}{\mathrm{\&Phi;}}_i^{*\left(k\right)}={\mathrm{\&omega;}}_i^{*2}M{\mathrm{\&Phi;}}_i^{*\left(k\right)}$ Calculate

The 3rd step: utilized for the 2nd step calculated

Compute matrix C and vectorial b are at last by α=(C ^TC) ^-1C ^TThe b calculation of alpha ^(k)

The 4th step: set stopping criterion for iteration, if max{| is α ^(k)-α ^(k-1)|≤tol sets up, iteration stopping then, and wherein tol is predefined allowable error; Otherwise returning for the 2nd step continues.

In superincumbent the finding the solution, the position of setting the damage member is known, when the position of damage member is unknown, must seek the damage position that a kind of method is confirmed structure earlier.The present invention proposes to adopt the variation of many class frequencys to damage the location.That is: have only real damage position just can cause the variation of the many class frequencys that conform to fully with actual test.For this reason; Damage position to each " setting "; Adopt above-mentioned method to carry out the degree of injury assessment, the result is designated as

i=1 ... Nm; Wherein subscript k representes k the damage position of setting, and subscript i representes to adopt the i class frequency to change to carry out the degree of injury assessment, and Nm representes available frequencies set of variations number.

To real damage position; When utilizing the variation of Nm class frequency to carry out the degree of injury assessment, the estimated value that obtains in theory should be to equate fully.But because the error of calculation or practical frequency error; Estimated value

can be inequal fully, but should approximately equal.So introduce the size that the coefficient of variation is represented its difference, be defined as

$e_k=\frac{{\mathrm{\&sigma;}}_k}{\stackrel{\&OverBar;}{{\mathrm{\&alpha;}}_k^i}},i=1,\&CenterDot;\&CenterDot;\&CenterDot;,\mathrm{Nm}$

In the following formula,

The average of representing the Nm group estimated value of corresponding k " setting " damage operating mode; σ _kBe its standard deviation.So e _kMiddle minimum value corresponding cells is located the factor for damage unit, definition damage

${\mathrm{DI}}_k=\frac{\min \left(e_k\right)}{e_k}$

Wherein, min (e _k) represent e _kGet minimum value; Then to equal 1 unit be real damage position to the damage location factor.After damage position was confirmed, the order of severity of damage can be confirmed through aforesaid method.

In above-mentioned second step, the m rank model frequency of identification of damage structure

concrete steps are following:

(1), utilize the structural dynamic response data after the sensor measurement works damage, and data storage is gone in the storer;

(2), from storer, read stored parameters information, utilize its limited lower mode frequency of Modal Parameters Identification identification as damaged structure model frequency

Two, set up plane steelframe finite element model

As shown in Figure 1, the plane steel frame construction of this example modeling effort is made up of members such as vertical bar, horizon bar, braces, totally 21 unit.Wherein, vertically bar unit is (1,5,9,13,17,21), and the horizontal bars unit is (2,6,10,14,18), and the brace unit is (3,7,11,15,19).Utilize MATLAB software programming finite element program, set up finite element model, as the benchmark finite element model of damaged structure not through computing machine.Then, simulate different damage operating modes again, draw the lower mode frequency of simulation actual measurement.This example has been simulated five kinds of damage operating modes, comprises the single component damage of diverse location, two component damages and damage in various degree.Damage position (as shown in Figure 2).

Three, damage location and degree of injury analysis and assessment

Based on first three rank practical frequency, utilize method of the present invention to damage location and degree of injury assessment, each operating mode is described as follows:

Damage operating mode one, horizontal bars unit 6, element stiffness loss 20%; Can damage the location exactly with the inventive method; The damage location factor (as shown in Figure 3), pointing out to damage member exactly is horizontal bars unit 6, the degree of injury estimated value is 20.007%.

Damage operating mode two, vertical bar element 9, element stiffness loss 10%; Can damage the location exactly with the inventive method; The damage location factor (as shown in Figure 4), pointing out to damage member exactly is vertical bar element 9, the degree of injury estimated value is 10.005%.

Damage operating mode three, brace unit 11, element stiffness loss 15% can damage the location exactly with the inventive method, the damage location factor (as shown in Figure 5), pointing out to damage member exactly is brace unit 11, the degree of injury estimated value is 15.038%.

Damage operating mode four; Horizontal bars unit 6, brace unit 11 damage simultaneously; Its rigidity loses 20% and 10% respectively, can damage the location exactly with the inventive method, the damage location factor (as shown in Figure 5); It is horizontal bars unit 6 that its x axle points out to damage member exactly, and it is brace unit 11 that the y axle points out to damage member exactly; Level of damage bar element 6 is respectively 19.992% and 10.004% with the degree of injury estimated value of damage brace unit 11.Verified this method to the damage of two places takes place, that is: horizontal bars and brace can well damage the assessment of location and degree of injury.

Damage operating mode five; Brace unit 7 damages with vertical bar unit 13 simultaneously; Its loss of rigidity is respectively 20% and 10%, can damage the location exactly with the inventive method, the damage location factor (as shown in Figure 6); It is brace unit 7 that its x axle points out to damage member exactly, and it is vertical bar unit 13 that the y axle points out to damage member exactly; The degree of injury estimated value of damage brace unit 7 and the vertical bar unit 13 of damage is respectively 19.992% and 10.004%.Verified that this method can well damage the location and assess with degree of injury two places damage (brace unit 7 with vertically bar unit 13) takes place.

The above; It only is preferred embodiment of the present invention; Be not that the present invention is done any pro forma restriction, every foundation technical spirit of the present invention all still belongs in the scope of technical scheme of the present invention any simple modification, equivalent variations and modification that above embodiment did.

Claims (6)

1. one kind based on the appraisal procedure of the structural damage of change of frequency location with degree of injury, it is characterized in that: adopt following steps:

The first step: set up the not finite element model of damaged structure, as the benchmark finite element model;

Second step: damaged structure is carried out vibration-testing, the m rank model frequency of identification of damage structure ${\mathrm{\&omega;}}_i^{*}(i=1,...,m);$

The 3rd step: damaged structure is carried out the degree of injury assessment;

The 4th step: damaged structure is carried out damage position confirm.

2. according to claim 1 based on the structural damage location of change of frequency and the appraisal procedure of degree of injury; It is characterized in that: in the said first step; The benchmark finite element model representes with K and M, that is: the not stiffness matrix of damaged structure and mass matrix, and its list of feature values is shown

$K{\mathrm{\&Phi;}}_i={\mathrm{\&omega;}}_i^{2}M{\mathrm{\&Phi;}}_i---\left(1\right)$

Wherein, ω _iAnd Φ _iBe respectively the i order frequency and the vibration shape of benchmark finite element model.

3. according to claim 1 based on the structural damage location of change of frequency and the appraisal procedure of degree of injury; It is characterized in that: in said second step, the m rank model frequency of identification of damage structure

concrete steps are following:

(1), utilize the structural dynamic response data after the sensor measurement works damage, and data storage is gone in the storer;

(2), from storer, read stored parameters information, utilize its limited lower mode frequency of Modal Parameters Identification identification as damaged structure model frequency

4. according to claim 1,2 or 3 described structural damages location based on the change of frequency appraisal procedure with degree of injury, it is characterized in that: said damaged structure is with the not stiffness matrix and the mass matrix eigenwert formula of damaged structure:

Be benchmark, with K ^*And M ^*The integral rigidity matrix and the total quality matrix of expression damaged structure, structural damage generally only causes the variation of the rigidity of structure, and very little to the quality influence of structure, sets M ^*=M, then its eigenwert does

$K^{*}{\mathrm{\&Phi;}}_i^{*}={\mathrm{\&omega;}}_i^{*2}M{\mathrm{\&Phi;}}_i^{*}---\left(2\right)$

Wherein, and

is respectively the i order frequency and the vibration shape of damaged structure; Utilize formula (1) and (2), obtain

${\left({\mathrm{\&Phi;}}_i\right)}^T{K}^{*}{\mathrm{\&Phi;}}_i^{*}=\frac{{\mathrm{\&omega;}}_i^{*2}}{{\mathrm{\&omega;}}_i^2}{\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*}---\left(3\right)$

Wherein, subscript T represents transpose of a matrix, and promptly ranks exchange;

Setting structure has N _dDamage has taken place in individual unit, and damage position is known, and then the integral rigidity matrix representation of damaged structure does

$K^{*}=K+\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{K}_{l_n}---\left(4\right)$

Wherein, α _nAnd l _nBe respectively the degree of injury of n damage unit and the unit number of n damage unit; With formula (4) substitution formula (3), can get

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2}{\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*}---\left(5\right)$

Definition $C_i={\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*},$ $C_{n,i}={\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}$ With $b_i=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2},$ Then formula (5) can be written as

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{c}_{n,i}=b_i---\left(6\right)$

Write as matrix form, promptly

Cα＝b (7)

Set the preceding m order frequency information that has recorded damaged structure, then C is that dimension is m * N _dMatrix, α and b are respectively that dimension is N _d* 1 and the column vector of m * 1; Through least square method solution formula (7), obtain

α＝(C ^TC) ^-1C ^Tb (8)

α is the degree of injury of structural unit.

5. according to claim 1 based on the structural damage location of change of frequency and the appraisal procedure of degree of injury, it is characterized in that: in said the 3rd step, the concrete steps of damaged structure being carried out the degree of injury assessment are following:

(1) the setting damage position is known; Utilize the model frequency

of benchmark finite element model and damaged structure to make up m equation, that is:

$\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n{\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2}{\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*}$

Wherein, K is the global stiffness matrix of benchmark finite element model;

And α _nBe respectively the element stiffness matrix and the degree of injury of n damage unit; N _dNumber for the damage unit; Φ _iWith

Be respectively the i first order mode of benchmark finite element model and damaged structure; ω _iWith Be respectively the i order frequency of benchmark finite element model and damaged structure; Subscript T represents transpose of a matrix;

Definition $C_i={\left({\mathrm{\&Phi;}}_i\right)}^{T}K{\mathrm{\&Phi;}}_i^{*},$ $C_{n,i}={\left({\mathrm{\&Phi;}}_i\right)}^T{K}_{l_n}{\mathrm{\&Phi;}}_i^{*}$ With $b_i=\frac{{\mathrm{\&omega;}}_i^{*2}-{\mathrm{\&omega;}}_i^2}{{\mathrm{\&omega;}}_i^2},$ Then following formula can be written as matrix form C α=b;

(2) iterative degree of injury;

The concrete calculation procedure of iterative degree of injury is:

1. iteration initial assignment, k=0, setting degree of injury is 0, i.e. α ⁽⁰⁾=0, subscript " 0 " is represented iterative initial value;

2. iteration begins, k=k+1, by $K^{*\left(k\right)}=K+\underset{n=1}{\overset{N_d}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_n^{(k-1)}{K}_{l_n}$ And formula $K^{*\left(k\right)}{\mathrm{\&Phi;}}_i^{*\left(k\right)}={\mathrm{\&omega;}}_i^{*2}M{\mathrm{\&Phi;}}_i^{*\left(k\right)}$ Calculate

3. 2. utilization calculates

Compute matrix C and vectorial b, last, by α=(C ^TC) ^-1C ^TThe b calculation of alpha ^(k)

4. set stopping criterion for iteration, if max{| is α ^(k)-α ^(k-1)|≤tol sets up, iteration stopping then, and wherein, tol is predefined allowable error, then α ^(k)Be the degree of injury estimated value; Continue otherwise return 2..

6. according to claim 1 based on the structural damage location of change of frequency and the appraisal procedure of degree of injury, it is characterized in that: in said the 4th step, the concrete calculation procedure that damage position is confirmed is:

(1) sets total N _kIndividual damage operating mode is utilized N _mClass frequency changes carries out the degree of injury assessment to each damage operating mode, obtains N _mIndividual degree of injury estimated value

I=1 ..., N _m, subscript k representes k the damage operating mode of setting;

(2) utilize the degree of injury estimated value

Make up damage location factor D I _k, it is defined as

${\mathrm{DI}}_k=\frac{\min \left(e_k\right)}{e_k}$

In the following formula

Represent corresponding k N that sets the damage operating mode _mThe average of group estimated value, σ _kBe its standard deviation, min (e _k) represent e _kGet minimum value; Then to equal 1 unit be real damage position to the damage location factor.

Images